• Asia pacific journal of information systems
• /
• v.19 no.2
• /
• pp.157-178
• /
• 2009

Corporate credit rating is a very important factor in the market for corporate debt. Information concerning corporate operations is often disseminated to market participants through the changes in credit ratings that are published by professional rating agencies, such as Standard and Poor's (S&P) and Moody's Investor Service. Since these agencies generally require a large fee for the service, and the periodically provided ratings sometimes do not reflect the default risk of the company at the time, it may be advantageous for bond-market participants to be able to classify credit ratings before the agencies actually publish them. As a result, it is very important for companies (especially, financial companies) to develop a proper model of credit rating. From a technical perspective, the credit rating constitutes a typical, multiclass, classification problem because rating agencies generally have ten or more categories of ratings. For example, S&P's ratings range from AAA for the highest-quality bonds to D for the lowest-quality bonds. The professional rating agencies emphasize the importance of analysts' subjective judgments in the determination of credit ratings. However, in practice, a mathematical model that uses the financial variables of companies plays an important role in determining credit ratings, since it is convenient to apply and cost efficient. These financial variables include the ratios that represent a company's leverage status, liquidity status, and profitability status. Several statistical and artificial intelligence (AI) techniques have been applied as tools for predicting credit ratings. Among them, artificial neural networks are most prevalent in the area of finance because of their broad applicability to many business problems and their preeminent ability to adapt. However, artificial neural networks also have many defects, including the difficulty in determining the values of the control parameters and the number of processing elements in the layer as well as the risk of over-fitting. Of late, because of their robustness and high accuracy, support vector machines (SVMs) have become popular as a solution for problems with generating accurate prediction. An SVM's solution may be globally optimal because SVMs seek to minimize structural risk. On the other hand, artificial neural network models may tend to find locally optimal solutions because they seek to minimize empirical risk. In addition, no parameters need to be tuned in SVMs, barring the upper bound for non-separable cases in linear SVMs. Since SVMs were originally devised for binary classification, however they are not intrinsically geared for multiclass classifications as in credit ratings. Thus, researchers have tried to extend the original SVM to multiclass classification. Hitherto, a variety of techniques to extend standard SVMs to multiclass SVMs (MSVMs) has been proposed in the literature Only a few types of MSVM are, however, tested using prior studies that apply MSVMs to credit ratings studies. In this study, we examined six different techniques of MSVMs: (1) One-Against-One, (2) One-Against-AIL (3) DAGSVM, (4) ECOC, (5) Method of Weston and Watkins, and (6) Method of Crammer and Singer. In addition, we examined the prediction accuracy of some modified version of conventional MSVM techniques. To find the most appropriate technique of MSVMs for corporate bond rating, we applied all the techniques of MSVMs to a real-world case of credit rating in Korea. The best application is in corporate bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. For our study the research data were collected from National Information and Credit Evaluation, Inc., a major bond-rating company in Korea. The data set is comprised of the bond-ratings for the year 2002 and various financial variables for 1,295 companies from the manufacturing industry in Korea. We compared the results of these techniques with one another, and with those of traditional methods for credit ratings, such as multiple discriminant analysis (MDA), multinomial logistic regression (MLOGIT), and artificial neural networks (ANNs). As a result, we found that DAGSVM with an ordered list was the best approach for the prediction of bond rating. In addition, we found that the modified version of ECOC approach can yield higher prediction accuracy for the cases showing clear patterns.

• Journal of Bio-Environment Control
• /
• v.7 no.1
• /
• pp.15-24
• /
• 1998

The greenhouse temperature controls in general have been managed by the above-ground part environment, But the temperature of root zone was known very important factor for the 9rofth and the yield of vegetables in greenhouse. The purpose of this study is to develop a good method for cultivation using solar energy which can apply warming soil and to develop the greenhouse soil temperature automatic control system. Followings are summary of this study：1 When the greenhouse inner temperature changes were about 24$^{\circ}C$ during a day in October, the temperature of non-warmed soil was differenced 6$^{\circ}C$ in the depth 10cm and 3$^{\circ}C$ in the depth 20cm. 2. When water supply temperature was kept at 40, 50 and 6$0^{\circ}C$, the lowest soil temperature in the depth of 10cm is 2$0^{\circ}C$ and that of 20cm was 23$^{\circ}C$. and when the water supply temperature was over 4$0^{\circ}C$, the space heating temperature did not affect the temperature variation of soil. 3. In comparison with conditions of the warmed and non-warmed soil, when the water supply temperature is 28$^{\circ}C$, soil temperatures had the high temperature of 4$0^{\circ}C$~7$^{\circ}C$ in the depth of 10cm to 20 cm. 4. The line of boundary area was appeared in the depth of 15~20cm, 13~19cm and 12~17cm. when the water supply temperature was 4$0^{\circ}C$, 5$0^{\circ}C$ and 6$0^{\circ}C$. 5. When th inner greenhouse air temperature is maintained over 11$^{\circ}C$ and the water supply temperature is supported 28$^{\circ}C$, the lowest temperature is kept up over 2$0^{\circ}C$.

• Journal of the Mineralogical Society of Korea
• /
• v.29 no.3
• /
• pp.155-165
• /
• 2016

Understanding the effect of boron content on atomic structures of boron-bearing multicomponent silicate melts is essential to reveal the atomistic origins of diverse geochemical processes involving silica-rich magmas, such as explosive volcanic eruption. The detailed atomic environments around B and Al in boron-bearing complex aluminosilicate glasses yield atomistic insights into reactivity of nuclear waste glasses in contact with aqueous solutions. We report experimental results on the effect of boron content on the atomic structures of sodium borate glasses and boron-bearing multicomponent silicate melts [malinkoite ($NaBSiO_4$)-nepheline ($NaAlSiO_4$) pseudo-binary glasses] using the high-resolution solid-state NMR ($^{11}B$ and $^{27}Al$). The $^{11}B$ MAS NMR spectra of sodium borate glasses show that three-coodrinated boron ($^{[3]}B$) increases with increasing $B_2O_3$ content. While the spectra imply that the fraction of non-ring species decreases with decreasing boron content, peak position of the species is expected to vary with Na content. Therefore, the quantitative estimation of the fractions of the ring/non-ring species remains to be explored. The $^{11}B$ MAS NMR spectra of the glasses in the malinkoite-nepheline join show that four-coordinated boron ($^{[4]}B$) increases as $X_{Ma}$ [$=NaBSiO_4/(NaBSiO_4+NaAlSiO_4)$] increases while $^{[3]}B$ decreases. $^{27}Al$ MAS NMR spectra of the multicomponent glasses confirm that four-coordinated aluminum ($^{[4]}Al$) is dominant. It is also observed that a drastic decrease in the peak widths (full-width at half-maximum, FWHM) of $^{[4]}Al$ with an addition of boron ($X_{Ma}=0.25$) in nepheline glasses. This indicates a decrease in structural and topological disorder around $^{[4]}Al$ in the glasses with increasing boron content. The quantitative atomic environments around boron of both binary and multicomponent glasses were estimated from the simulation results of $^{11}B$ MAS NMR spectra, revealing complex-nonlinear variation of boron topology with varying composition. The current results can be potentially used to account for the structural origins of the change in macroscopic properties of boron-bearing oxide melts with varying boron content.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.32 no.7
• /
• pp.1039-1045
• /
• 2003

Functional healthy drinks were processed with freeze dried powders of ethanol extract from of defatted safflower (Carthamus tinctorious L.) seed cake and some useful components of the drinks were investigated. Yield of freeze dried powder was the highest as 8.42% when it extracted with 60% ethanol (60% EFDP). Each drink contained 0.02% of freeze dried powder and ranged 10.6∼13.8% of soluble solid, 2.90∼3.68 of pH, 0.10∼0.83% of titratable acidity. ‘L’ value of drink-I (DSD-I) was the highest as 94.82$\pm$2.45, ‘b’ and ‘a’ value of drink-V (DSD-V) was highest as 27.15-2.65 and 28.67$\pm$2.69, respectively. Major free sugars of drink were 6015.3∼7918.2 mg% of glucose and 1511.4∼2091.0 mg% of sucrose. The content of citric acid was the highest as 179.2∼981.3 mg%. The content of total phenol in 60% EFDP was 99.17 mg% and that of drink-II(DSD-II) and DSD-V was 307.84 mg% and 224.06 mg%, respectively. Total flavonoid was contained as 50.29 mg% in 80% ethanol extract (80% EFDP) and 125.20 mg% in DSD-V. N-[2-(5-hydroxy-1H-indol-3-yl) ethyl] ferulamide (serotonin-I) was determined as high as 18.81 ppm in 80% EFDP and ranged 2.42∼2.89 ppm in drinks. N-[2-(5-hydroxy-lH-indol-3yl)ethyl]-p-coumaramide (serotonin-II) was determined as 30.17 ppm in 80% EFDP and ranged 3.79∼4.59 ppm in drinks. Acacetin, flavonoid compound were 9.83 ppm in amyloglucosidase hydrosis + 60% ethanol extract (A + 60% EFDP) and ranged 0.98∼1.26 ppm in drinks. Electron donating ability (EDA, %) was measured and compared with 100 ppm BHA as chemical antioxidant. EDA was 93.97$\pm$2.21% in A+60% EFDP, 94.79$\pm$2.26% in DSD-I, 94.69$\pm$1.37% in DSD-II, and 93.83$\pm$1.49% in BHA. DSD-II added with hot water extract solution from Korean ginseng and safflower yellow pigment recorded the highest sensory score.

• Journal of the Korean Electrochemical Society
• /
• v.6 no.1
• /
• pp.59-64
• /
• 2003

We fabricated mixed ionic-electronic conducting membranes, $CH_4\;Using\;{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$, by solid state reaction method for solid oxide fuel cell. The membranes consisted of single perovskite phase and exhibited high relative density, $>95\%$. We coated $La_{0.6}Sr_{0.4}CoO_{3-\delta}$ layer using screen printing method in order to improve surface reactivity of the $La_{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$. As a result, the oxygen permeation flux of the coated $La_{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$ showed higher value, $0.5ml/min{\cdot}cm^2\;at\;950^{\circ}C$ than the uncoated one. Higher oxygen permeation was observed in the porously coated Lao $La_{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$membranes with larger grain sizes. Syngas, $CO+H_2$, was successfully obtained from methane gas, $CH_4$, using the $La_{0.6}Sr_{0.4}CoO_{3-\delta}$ coated $La_{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$, with over $40\%\;of\;CH_4$ conversion and syngas yield. $La_{0.7}Sr_{0.3}Ga_{0.6}Fe_{0.4}O_{3-\delta}$ membrane was stable even when it was exposed to the reducing environment, methane, for 600 hrs at $950^{\circ}C$.

• Korean Journal of Environmental Agriculture
• /
• v.15 no.3
• /
• pp.362-371
• /
• 1996

This study was conducted to investigate the biochemical properties of isolated bacteria, low temperature tolerant methanogens which were selected for use as inoculum for anaerobic fermentation of agricultural and livestock wasted at low temperature. The results, obtained were summarized as follows: Low temperature tolerant methanogens were isolated from the samples which showed the high methanogenesis rate by enrichment culture at low temperature in methanol medium. These methanogens, Methanobacterium M-251 and Methanobacterium M-253 were isolated from swampy sediment at latitude $56.9^{\circ}$, Methanosarcina mazei M-372 from lake sediment IV at latitude $55.0^{\circ}$ N, and Methanobacterium formicicum M-375 from tidal land soil at latitude $37.0^{\circ}N$, respectively. The isolated anaerobic bacteria could not use sugars as carbon sources. The optimum pH value for the growth of M-251 and M-375 was 6.8, but those for M-253 and M-372 6.5 and 7.0, respectively. The minimum growth temperature of isolated, M-251 and M-253 were $8^{\circ}C$ and the optimum temperature $30^{\circ}C$, while the minimum of M-392 and M-395 were $13^{\circ}C$ and the optimum $37^{\circ}C$. The growth rate of isolates at $17.5^{\circ}C$ were lower by 32-50% than that of $30^{\circ}C$. The isolated Methanobacterium strains such as M-251, M-253, and M-375 have lower cell yield, 0.38-1.21g/1M $CH_4$ than 1.14-1.51g/1M $CH_4$ of Methanosarcina mazei M-372.

• Journal of Animal Science and Technology
• /
• v.44 no.6
• /
• pp.793-800
• /
• 2002

The characteristics of carcass and meat yields of fattening pigs by production steps were investigated with Landrace (LL, n=41), Yorkshire (YY, n=33), and Duroc (DD, n=30), $F_1$ (LY, n=25), the Crossbred of LYD (n=48). Duroc had more weight loss in carcasses weight than that of the other breed(p<0.05). Yorkshire and $F_1$ had higher production weight in retail cut than that of the other pure or crossbred. Carcass from Landrace and $F_1$ were significantly longer in length than the other breed(p<0.05). $F_1$ produced wider carcass than the other breed. Carcass thickness at aitch bone was higher for $F_1$ and Crossbred than the other breed(p<0.05). Landrace, Yorkshire and $F_1$ produced more loin and tenderloin in weight than the other breed (p<0.05). Yorkshire and $F_1$ produced more picnic shoulder when compared to the other breed. The hind legs produced from Yorkshire and $F_1$ were higher in weight and the fore legs produced from the Crossbred were higher in weight. Duroc produced the lowest weight of belly among the breeds. The acceptance level of loin were extremely low for all breed. Landrace had the highest acceptance level for tenderloin. Yorkshire had the highest acceptance level for picnic shoulder and ham when evaluated by export standard of Japan. In conclusion, The introduction of pure breed and establishment of mating steps are necessary to produce highly accepted pork with high acceptance in carcass and meat yields.

• Journal of Animal Science and Technology
• /
• v.52 no.5
• /
• pp.427-434
• /
• 2010

This study was conducted to investigate the growth performance and meat quality improvement according to castration, optimal feeding management and ruminally protected amino acid-enriched fatty acid (RPAAFA) for the unselected Hanwoo bulls in the performance test. Bulls were castrated at approximately 14 months of age. Sixteen Hanwoo steers, 15 months of age and weighing $412.9{\pm}24.9kg$, were distributed into 2 groups. Steers were fed a basal diet supplemented with RPAAFA at 0 g (control) or 100 g (treatment), respectively for 12 months. Steers were slaughtered at 27 months of age. Average daily gain for treatment tended to be higher (p=0.10) than that of control, whereas feed conversion ratio tended to be lower (p=0.07) in treatment than in control. The supplementation of RPAAFA did not affect rib eye area, back fat thickness, meat color, fat color, texture and maturity. The appearance rates of yield 'A' grade and high quality grade ($1^{++}$, $1^+$ and 1) were higher in treatment than in control. The content of moisture, fat, protein and ash in longissimus muscles were similar between control and treatment. The supplementation of RPAAFA did not affect water-holding capacity, oxidation and reduction potential, myoglobin and fatty acid contents in longissimus muscles. Thus, present results indicate that castration, optimal feeding management and RPAAFA may be recommended for improving growth performance and quality grade of the unselected Hanwoo bulls in the performance test.

• Journal of the Korea Organic Resources Recycling Association
• /
• v.9 no.1
• /
• pp.65-72
• /
• 2001

Four newly isolated bacteria from soil were used to manufacture microbial inoculum to compost food waste. The bacteria, GM103, V25, V31, and V35, were identified as Bacillus licheniformis, B. subtilis, B. stearothermophilius, and B, subtilis, respectively. The bacterial strains were efficient to degrade protein and starch and also able to inhibit the growth of plant pathogenic fungus Rhizopus stronifer. The GM103 showed distinct capability in degrading starch, but grow only aerobically. The other three bacterial strains. V25, V31, and V35, could grow both aerobically as well as anaerobically, in 10%(w/v) salt, at $50^{\circ}C$, and had good viability and survival rate in soil. These characteristics of the bacterial strains are very adquate in Korean food composting containing high concentration of salt, especially at home. By mixing the 4 bacterial culture broth with molasses, beet pulp, zeolite, The bacterial inoculum for food waste composting-BIOTOP-CLEAN-was made. The performance of food waste composting by the BIOTOP-CLEAN was compared with that by control(not treated) and HS(other demestic company's inoculum product for food waste composting). The maximum temperature of the food waste during the composting with the BIOTOP-CLEAN was $50^{\circ}C$, while those of the control and HS were $30^{\circ}C$ and $35^{\circ}C$, respectively. The BIOTOP-CLEAN gave the good smell and showed dark brown color, while the control gave bad smell and HS gave less bad smell. These indicates that the food waste composting by the BIOTOP-CLEAN had been well accomplished. The culture broth of V25, V31, V35 were sparyed to the plants of tomato, chinese cabbage, raddish, red pepper every month and the spraying the culture broth to these plant significantly improved the production yield of the crops, due to the control effect of the bacterial strains against the plant pathogens.

• Journal of Mushroom
• /
• v.18 no.1
• /
• pp.1-9
• /
• 2020

Mycorrhiza refers to the association between a plant and a fungus colonizing the cortical tissue of the plant's roots during periods of active plant growth. The benefits afforded by plants from mycorrhizal symbioses can be characterized either agronomically, based on increased growth and yield, or ecologically, based on improved fitness (i.e., reproductive ability). In either case, the benefit accrues primarily because mycorrhizal fungi form a critical linkage between plant roots and the soil. The soilborne or extramatrical hyphae take up nutrients from the soil solution and transport them to the root. This mycorrhizae-mediated mechanism increases the effective absorptive surface area of the plant. There are seven major types of mycorrhizae along with mycoheterotrophy: endomycorrhizae (arbuscular mycorrhizae, AM), ectomycorrhizae (EM), ectendomycorrhizae, monotropoid, arbutoid, orchid, and ericoid. Endomycorrhizal fungi form arbuscules or highly branched structures within root cortical cells, giving rise to arbuscular mycorrhiza, which may produce extensive extramatrical hyphae and significantly increase phosphorus inflow rates in the plants they colonize. Ectomycorrhizal fungi may produce large quantities of hyphae on the root and in the soil; these hyphae play a role in absorption and translocation of inorganic nutrients and water, and also release nutrients from litter layers by producing enzymes involved in mineralization of organic matters. Over 4,000 fungal species, primarily belonging to Basidiomycotina and to a lesser extent Ascomycotina, are able to form ectomycorrhizae. Many of these fungi produce various mushrooms on the forest floor that are traded at a high price. In this paper, we discuss the benefits, nutrient cycles, and artificial cultivation of mycorrhizae in Korea.